1. Atmospheric neutrino fluxes Teresa Montaruli, Paolo Desiati, Aya Ishihara, UW, IceCube Meeting, Mar 2005 A background and an interesting measurement How well do we know the atmospheric neutrino fluxes? The primary CR spectrum Can we have a muon and neutrino flux from the same code? A tool for simulation and analysis: a C++ class and a library for fluxes (suitable for ANIS and IceTray) It is a beautiful measurement with the largest data sample AMANDA (IceCube) have since K physics and prompt Fluxes are poorly constrained by accelerator data

2. Super-Kamiokande and MACRO MACRO final data Bartol 96 HKKM (2001 CR) FLUKA (2001 CR) FLUKA (1996 CR) E  >1 GeV predictions no oscill. ~20%  Honda 2001  Honda 1995  Bartol 1996  SK Data predictions  m 2 =2.5 10 -3 eV 2 SK and MACRO through-going muon data pointed out that the CR assumed after ICRC2001 by HKKM and Bartol was too steep and that a better agreement was achieved with the previous Agrawal et al. 1996 CR flux Moreover ATIC data preferred E -2.7 to steeper spectra TM, ICRC2003 HE rapporteur talk

3. Primary CRs Primary spectra: ICRC2001 (softer), Agrawal or Old Bartol (harder) He p 5% agreement between AMS and Bess98 < 100 GeV BUT CAPRICE -20% ICRC2001 assumes AMS-Bess The region relevant for AMANDA/IceCube is higly uncertain due to large errors of Jacee and RunJob data Moreover He and heavy components are worse known and their relevance increases

4. The inputs: primary CRs in most recent calculations Main difference between ICRC2001/Gaisser-Honda 2002: CNO+Si-Mg+Fe E -2.7 /harder Main difference between ICRC2001/HKKM 2004: He [average of E -2.74 +E -2.64 ]/ E -2.64 ] + protons [E -2.74 / E-2.74+E -2.71 (E>100 GeV)] Results: HKKM 2004 is closer to previous Agrawal et al, 1996, harder than ICRC2001 and Gaisser-Honda fit resulting in more through-going muons Gaisser-Honda 2002

5. Theoretical Errors Though HKKM 2004 uses a harder spectrum than Bartol, it is lower! Interaction models are different (Target 2/DPMJET-III – to be updated 15% lower than L3  data) Lots of comparisons between Bartol, HKKM and FLUKA groups resulted in uncertainty reduction also thanks to muon data in the atmosphere. These groups published their results in 2003-2004. Can AMANDA data help? How does Lipari flux (used in AMANDA) compare to the more recent calculations? These results agree with: Gaisser astro-ph/0502380

6. hep-ex/0501064 AMANDA works in a higher E region where oscillations play a small role BUT in the region where hadronic models are worsely known (K physics, prompts…) Systematics need to be reduced – Paolo’s talk!!!! AMANDA DATA SK data prefer higher fluxes by 12% and harder spectra than HKKM2004 +0.05 >100 GeV

7. Atmospheric s and  s with FLUKA FLUKA (http://www.fluka.org) new runs extending up to 10 6 GeV/nucleon started at UW. The goal is to have neutrinos from 4  and atmospheric muons from the same interaction and transport code (first 3D code, Astr.Phys.19, 2003)http://www.fluka.org Prompt neutrinos and muons will be addressed when DPMJET-III interface will be ready FLUKA yet poor statistics L3  data can help reducing systm at HE

8. The inputs: primary CR for VHE spectrum NEW FLUKA generation: CR suitable for VHE

9. The AtmosphericFlux Class The constructor: AtmosphericFlux mymodel(model); Class Methods: Init: Read Tables in prepared in common format Flux: returns the flux at a given energy, angle and for a flavor eg:double flux = mymodel.Flux(1,1.e3,0.1); returns: flux in GeV -1 cm -2 sr -1 s -1 given: Neutrino type: e, anti- e, , anti-  Neutrino Energy Cos(zenith) Average flux: returns averaged over lower hemisphere Compatible with ANIS Usable by IceTray (A. Hishihara) Usable in analysis stage with ROOT as a shareable library Extensible to ANY FLUX (can become a library of fluxes to be used for event reweighting purposes)